CHEMICAL SCIENCE. 229 



water, diffusion takes place in quantity which has a relation" to the per- 

 centage of salt dissolved. Alcohol, however, exhibits an anomaly in this 

 respect ; for the quantity of alcohol which diffused itself through the 

 membrane, when 5 per cent, of alcohol was present in the liquid, was not 

 increased when the percentage of alcohol was 10, 15, or 20. The phe- 

 nomenon indicates a sifting or separating power to reside in membrane, 

 and introduces a third element, in addition to diffusion and osmose, into 

 the discussion of the permeability of membranous septa. The author 

 believed that Sb'mmering's experiment was an instance of arrested diffusion 

 where more than o per cent, of alcohol was present. The action has some 

 resemblance to the separating and secreting power of cells in the living 

 organism, and may prove of great physiological interest, particularly if the 

 action should be found to extend to albumen and other organic substances. 

 Prof. Faraday considered the latter part of the paper exceedingly 

 important, and expressed a wish that Prof, graham would give his reasons 

 for believing that liquids diffused, owing to a repulsion between the liquid 

 particles. Might not the attraction of the surrounding medium be wholly 

 or partly the cause ? In answer to this, Prof. Graham stated that the phe- 

 nomena characteristic of gaseous diffusion might be explained by an 

 attractive as well as a repulsive force. In the diffusion of liquids, the 

 same analogies were observed, as also the same intensity of action. From 

 a bottle containing solution of alum the sulphate of potash goes off first, 

 and sulphate of alumina remains. Again : sulphurous acid and chloride 

 of sodium may be boiled together, and no hydrochloric acid is given off ; 

 but mix them in the diffusion vial, and hydrochloric acid is given off, 

 whilst sulphite of soda remains. Experiments on this subject are being 

 accumulated by the author, and he sees every reason to consider that, since 

 gaseous diffusion can be most clearly explained by the repulsive vie\v, 

 liquid diffusion, so analogous to it, should be likewise expressed. 



RESEARCHES ON CHEMICAL AFFINITY. 



Marguerite, of Paris, has been engaged in some new experiments touch- 

 ing the chemical affinities, and he revives the oft-debated question, which 

 consists in defining and explaining the manner of grouping which the ele- 

 ments of two salts, dissolved in the same liquid, will adopt. Every salt 

 being formed of an acid element and a basic element, it may be asked 

 whether, after the dissolution and the mixture of the elements, they retain 

 their original association, or, if the mixture occasions a double dissolution, 

 they are more or less completely exchanged. When one of these possible 

 combinations is insoluble, Berthollet's law announces it will be formed, 

 and experience shows it does form. If, for example, the experimenter dis- 

 solves separately nitrate of lime and sulphate of potassa, the mixture of 

 the two solutions will contain every thing that is necessary to form the 

 nitrate of potassa and the sulphate of lime ; and as the latter is very lit- 

 tle soluble, it will be precipitated in abandoning the other elements which, 



